IoT and ML-Based Personalised Healthcare System for Heart Patients

In this paper, a remote health monitoring system using IoT is considered. In this, a prototype was developed which keeps track of the pulse rate and body temperature of a heart patient. This prototype is to be used by the patient remotely. The sensor data is then transmitted to an IoT server using a Wi-Fi module and a microcontroller. Moreover, this live data is further recorded in a spreadsheet which can be shared with concerned authorities to track the patient’s health condition. The system also allows a notification to be sent out via SMS/email when the patient’s condition worsens. The second part involves the prediction of whether a patient is at risk of having a heart condition based on their medical records. It involves data analysis, feature engineering, and deploying machine learning algorithms to create a data model that predicts whether a patient is at risk with a high amount of accuracy. In addition, a hardware system is implemented and realised that could be used by hospitals to remotely track the health condition of their patients. We were also able to achieve the additional feature of notifying concerned authorities if an anomaly is detected. With the help of data analysis and machine learning tools, we were able to analyse and find the correlation between various health factors and the risk level of having a heart condition. Finally, we created a data model that was able to predict patients that are potentially at risk with a high accuracy rate that could be used to differentiate the level of importance of each factor. The final system that was developed during the training period was a prototype that was able to counter some of the existing challenges in the area of work we focused on. We were also able to identify newer challenges and suggest further areas of research that need to be addressed.